Skip to main content
hc
Submit
mcpSource-backedReview first Safety ✓ Privacy ✓

MCP Toolbox for Databases

Open source MCP server and tool framework from Google for connecting AI agents, IDEs, and applications to databases through prebuilt or custom database tools.

by Google·added 2026-06-05·
Claude CodeClaude Desktop
HarnessClaude CodeClaude Desktop
Review first — review before installing

Open the source and read safety notes before installing.

Safety notes

  • Database tools can expose sensitive schemas and data, and some custom tools can modify data if configured that way.
  • Prefer least-privilege accounts, read-only roles for exploration, query limits, and approved toolsets.
  • Review every `tools.yaml` statement, parameter, source, and toolset before giving it to an agent.
  • Avoid broad `execute_sql` access in production; expose structured, parameterized tools where possible.
  • Query loops can create high database load, unexpected cloud costs, lock contention, or noisy audit logs.

Privacy notes

  • Schema names, table names, query text, query results, connection metadata, and business data may pass through the MCP server and AI client.
  • Database passwords, IAM credentials, service account keys, and connection strings are secrets and should never be committed or pasted into prompts.
  • Logs, traces, and metrics can include tool names, query metadata, error messages, and database identifiers.

Prerequisites

  • Node.js and npx for the npm server path, or another documented Toolbox runtime.
  • Database credentials scoped to the exact database, schema, and permissions needed.
  • Environment variables or secret storage for database connection settings.
  • A reviewed prebuilt database target or `tools.yaml` file before exposing tools to an agent.
  • MCP client such as Claude Code, Codex, Gemini CLI, Google Antigravity, or another compatible host.

Schema details

Install type
cli
Troubleshooting
No
Source repository stats
Scope
Source repo
Collection metadata
Estimated setup
15 minutes
Difficulty
intermediate
Full copyable content
{
  "mcpServers": {
    "toolbox-postgres": {
      "command": "npx",
      "args": ["-y", "@toolbox-sdk/server", "--prebuilt=postgres", "--stdio"]
    }
  }
}

About this resource

Content

MCP Toolbox for Databases is an open source MCP server and framework for database-backed tools. It can expose prebuilt database tools for MCP clients or serve custom tools defined in a tools.yaml file. The project documents support for Google Cloud databases and many other systems, including PostgreSQL, MySQL, SQL Server, Oracle, MongoDB, Redis, Elasticsearch, CockroachDB, ClickHouse, Couchbase, Neo4j, Snowflake, and Trino.

Use it when an agent needs database context, schema exploration, or pre-approved database operations without hand-building an MCP server for each database.

Source Review

These sources were reviewed on 2026-06-05. Prefer the live documentation for supported database lists, prebuilt tool names, runtime options, authentication support, and configuration schema.

Features

  • Prebuilt MCP tools for common database exploration and querying workflows.
  • Custom database tools defined through tools.yaml.
  • Support for parameterized SQL tools, toolsets, prompts, and source definitions.
  • Integrated authentication patterns, connection pooling, and OpenTelemetry observability.
  • Compatibility with MCP clients such as Claude Code, Codex, Gemini CLI, Google Antigravity, and other MCP-capable hosts.
  • SDKs for application integration in Python, JavaScript/TypeScript, Go, and Java.

Installation

For a basic prebuilt PostgreSQL setup:

claude mcp add toolbox-postgres -- npx -y @toolbox-sdk/server --prebuilt=postgres --stdio

For JSON-based MCP clients:

{
  "mcpServers": {
    "toolbox-postgres": {
      "command": "npx",
      "args": ["-y", "@toolbox-sdk/server", "--prebuilt=postgres", "--stdio"]
    }
  }
}

For custom tools, create a reviewed tools.yaml and run the server with:

npx @toolbox-sdk/server --config tools.yaml

Use Cases

  • Let an agent inspect database schemas while generating data-aware application code.
  • Expose read-only SQL exploration in a development database.
  • Build production-ready, parameterized database tools instead of broad query access.
  • Add semantic search or structured data tools to an agent runtime.
  • Centralize database access patterns behind a reviewed MCP tool configuration.

Safety and Privacy

Database MCP servers need explicit guardrails. Use least-privilege credentials, prefer read-only accounts for exploration, limit exposed toolsets, and keep production write operations behind reviewed custom tools. Avoid giving an agent unbounded query access to sensitive or expensive databases.

Never commit database credentials, IAM secrets, service account keys, or connection strings. Treat schemas, row samples, query errors, and trace metadata as potentially sensitive.

Duplicate Check

Existing entries cover specific database MCP servers such as BigQuery, PostgreSQL, MongoDB, Snowflake, ClickHouse, Redis, and Qdrant. This entry covers Google's broader MCP Toolbox for Databases project and its reusable server/tool framework.

#databases#sql#google#mcp#data

Source citations

Signals

Loading live community signals…

More like this, weekly

A short, calm digest of reviewed Claude resources. Unsubscribe any time.